XML tags for the Planetfall Leaders

[Pfeffersack]

This threads purpose is to find better values for the various tags in the leaderhead.xml in Planetfall to give the leaders more personality. Of course the bahaviour of the orginal SMAC is the first source, but I think it might be worth to look beyond that, if we can improve balance, atmoshere, AI quality or just make use of new options Civ4 offers.

The first part is a kind of guide to explain what each value means - it's no complete, especially the second part about attitudes and diplomatic behaviour relies on some guesswork.


Part One of the XML-values: Basic Attitude, War behaviour, (Tech) Trading, Wonders, Espionage


<iWonderConstructRand>0</iWonderConstructRand>
---> Likelihood to initiate a wonder
+ LOUIS XIV.: 50
- MONTEZUMA, SITTING BULL: 0


<iBaseAttitude>0</iBaseAttitude>
---> General basic attitude towards other civs
+ GANDHI and ZARA YAQOB: 2
- many other (mainly the Warmongers): -1


<iBasePeaceWeight>0</iBasePeaceWeight>
---> How much value has peace for a civ?
+ GANDHI: 10
- BRENNUS,DE GAULLE,GENGHIS,NAPOLEON,... 0


<iPeaceWeightRand>0</iPeaceWeightRand>
---> some random number between 0 and this is added to the above
+/- FOR ALL LEADERS: 3


<iWarmongerRespect>0</iWarmongerRespect>
---> The higher the number, the better a civ gets along with other warmongers
+ many of the warlike leaders: 2
- the pacifists: 0


<iEspionageWeight>0</iEspionageWeight>
--> (unsure) might be the “value” of espionage yields compared to other things like profuction, commerce, etc.
+ STALIN: 150; CATHERINE,MAO: 130
- GANDHI: 50; MANSA,TOKU: 60


<iRefuseToTalkWarThreshold>0</iRefuseToTalkWarThreshold>
---> How long does the AI refuse talking after going to war? Doubled, if the AI started the war
+ SITTING BULL: 12
- DE GAULLE: 5


<iNoTechTradeThreshold>0</iNoTechTradeThreshold>
--> (unsure) likely % of all civs which must know a tech, that an AI is willing to trade this tech
+ TOKU 100
- MANSA 0


<iTechTradeKnownPercent>0</iTechTradeKnownPercent>
--> max. no. of tech trades in an entire game an AI leader will do with another leader before thinking that he/she becomes too advanced…
+ MANSA 20
- ALEXANDER, BISMARCK, BRENNUS, GENGHIS, HUAYNA, ISABELLA and others: 5


<iMaxGoldTradePercent>0</iMaxGoldTradePercent>
---> % of the total Gold in Treasury available for Trading/Diplomacy
+ LOUIS XIV: 30
- around 1/2 of the leaders: 5


<iMaxGoldPerTurnTradePercent>0</iMaxGoldPerTurnTradePercent>
---> % of the total GPT available for Trading/Diplomacy
+ LINCOLN, PACAL, SURYAVARMAN: 15
- all others: 10


<iMaxWarRand>0</iMaxWarRand>
---> 1 out of [number] chance per turn, that an AI starts to plan a war
+ MANSA MUSA, GANDHI: 400
- BOUDICA, GENGHIS, MONTEZUMA, RAGNAR, SHAKA: 50


<iMaxWarNearbyPowerRatio>0</iMaxWarNearbyPowerRatio>
---> Only if total war and nearby target: AI multiplies its own strength with [number]% to compare with possible targets
+ ISABELLA, MONTEZUMA, SITTING BULL: 130
- DE GAULLE, ELISABETH: 80


<iMaxWarDistantPowerRatio>0</iMaxWarDistantPowerRatio>
---> as above, but for non-nearby targets
+ ISABELLA, NAPOLEON: 100
- SITTING BULL: 0


<iMaxWarMinAdjacentLandPercent>0</iMaxWarMinAdjacentLandPercent>
---> One of three checks happening to see if total war vs. nearby civ happens. % of AI tiles adjacent to a civ must be greater then [number] to consider the civ as a target
+ ELISABETH, SITTING BULL, TOKUGAWA: 4
- ALEXANDER, ASOKA, BISMARCK, BOUDICCA, BRENNUS, GENGHIS,...: 0


<iLimitedWarRand>0</iLimitedWarRand>
---> 1 out of [number] chance per turn, that an AI starts to plan a limited war
+ ELISABETH, GANDHI, HATSCHEPSUT, LINCOLN, MANSA, PERICLES,...: 200
- ALEXANDER, GENGHIS, MONTEZUMA, NAPOLEON, RAGNAR: 40


<iLimitedWarPowerRatio>0</iLimitedWarPowerRatio>
---> Only if limited war and nearby target: AI multiplies its own strength with [number]% to compare with possible targets
+ ALEX,GENGHIS,LOUIS,MONTEZUMA: 130
- GANDHI,JOAO,JUSTINIAN. .: 80


<iDogpileWarRand>0</iDogpileWarRand>
---> 1 out of [number] chance per turn, that an AI starts to plan a dogpile war
+ LINCOLN: 150 (many others without a clear pattern 100)
- DE GAULLE: 20 (many others without a clear pattern 25)


<iMakePeaceRand>0</iMakePeaceRand>
--> 1 out of [number] basic chance per turn, that AI considers to try to make peace with one of it enemies (many others factors involved in final decision)
+ SITTING BULL 120
- ASOKA,CATHERINE,DE GAULLE,ELISABETH,GANDHI,LINCOLN.LOUIS,…: 10


<iDeclareWarTradeRand>0</iDeclareWarTradeRand>
--> 1 out of [number] chance per turn and other civ they have contact with, that the AI tries to bribe them into on of their wars (likely affecting only other AIs as potential war allies)
+ PACAL: 60 (a bit strange?)
- all others: 40


<iDemandRebukedSneakProb>0</iDemandRebukedSneakProb>
--> %chance that an AI demanding something from someone else without success will plan a later executed war on the rejector (100=DoW for sure!)
+ WASHINGTON: 100; CATHERINE,RAGNAR: 80
- many of the more peaceful leaders: 0


<iDemandRebukedWarProb>0</iDemandRebukedWarProb>
--> same situation as in last tag, but this time %chance that war is declared immediately (not used in game)
+ GENGHIS,ROOSEVELT,SULEIMAN: 50
- many of the more peaceful leaders: 0


<iRazeCityProb>10</iRazeCityProb>
---> % after capturing a city that city will be razed (other factors might cause razing even if this is 0!))
+ GENGHIS: 75
- around 1/2 (pazifists): 0


<iBuildUnitProb>50</iBuildUnitProb>
---> Base chance that city chooses unit as build (further modified)
+ MEHMED,NAPOLEON,RAGNAR,SHAKA: 40
- GANDHI: 15


<iBaseAttackOddsChange>0</iBaseAttackOddsChange>
<iAttackOddsChangeRand>16</iAttackOddsChangeRand>
--> both seem to have an impact of how the AI overestimates battle odds and therefore its likelihood to attack. The first is a fix number, then two times a random number between 0 and [2.number] is drawn and added to 1). This value gets added to the battle odds(?)
+ NAPOLEON,RAGNAR: 4 (base)
- many others: 0 (base)
rand. value is 8 for all civs! (except barbs, from which I have copied the tags)

 

[Pfeffersack]

Part Two of the XML-values: Diplomatic actions and it consequences

This section contains several blocks of tags, each with many single ones:

1. Effect of diplomatic (shared peace, same religion) and physical status (military power, close borders) of a civ on the diplomacy with others
2. AI tresholds for certain diplomatic actions
3. Some flavour stuff (favourite civic, religion and traits, production flavours)
4. Determining how often an AI chooses a certain diplomatic proposal
5. Determining how much time must have passed at least before a certain diplomatic proposal is made again to the same civ
6. How long the AI remembers certain past events
7. How strong the diplomatic impact of certain actions on attitude is at all
8. How likely the AI is to declare war based on it's attitude
9. Again some flavour stuff (favourite AI unit stratgey and terrain improvement)

I won't go as much into detail as with the first part - on reason is that Maniac link basically does this a lot better then I would be able to for most parts; the other is that a few parts (including the ones where Maniac linked guide also does not provide clear information) remain unsolved.

I don't get for example the relation of the 6th and 7th section. Ok, the first is about forgeting...but not all actions are listed in both sections - does it mean that anything not in 6. is permanent? But then I don't understand why there are also ones which are only listed in 6., but not in 7. (e.g. MEMORY_RECEIVED_TECH_FROM_ANY, likey that thing with "You have traded with our worst enemy!")
Also, I wonder if the values given in 6. are an absolut number or a 1 out of X chance per turn (I'm pretty sure that it is the latter, but the guide says the opposite)

Then their is an inconsistency with how the tags are used for determing the effects of having the same or a different religion (1st section):

iSameReligionAttitudeChange - Modifier to the attitude of the leader towards any leader with the same religion. GS: 1


iSameReligionAttitudeDivisor - Modifier to the attitude of the leader towards any leader with the same religion. GS: 10


iSameReligionAttitudeChangeLimit - The maximum change in attitude (positive or negative) which can occur as an effect of iSameReligionAttitudeChange. GS: 2 - 7


iDifferentReligionAttitudeChange - Modifier to the attitude of the leader towards any leader with a different religion. GS: 0 - -2


iDifferentReligionAttitudeDivisor - Not known. GS: -5


iDifferentReligionAttitudeChangeLimit - The maximum change in attitude (positive or negative) which can occur as an effect of iDifferentReligionAttitudeChange. GS: -1

The red values just don't make sense for me, but I checked the XML files - it does not seem to be a typo error, that are the actual values used in the files.

 

[Pfeffersack]

One way find good values for all the tags might be answering questions about how leader X should behave regarding Y, as with this information it isn't hard to code the behaviour into the tags (though there is no need for everyone to answer every question in line; just use it as guideline in your descriptions, if you like and feel like it might help you to provide feedback) Here is my list of questions I have come up so far:


Economic

- Tech Trading ...how interested in acquiring new tech and how willing to share tech in return, especially a new, secret tech no one has yet?
- How interested is he/she in secret projects? Or is the entire concept of "secret projects" rather something the leader is against?
- Generally, how good have the relations to be that a leaders trades something (might be ressoruces, tech, map)? Are there any items the leader will trade even with a worde attitude or not even, when relations are better?
- Relationship to money - saving for domestic purposes or using it for diplomatic/trade activities?
- How important is espionage (think of military intelligence, security measures vs. sabotage and/or offensive action)?
- How much of the economic capacity is he/she willing to invest in military?
- Are there any specially flavors for the leader (attention: this might be one key to solving the excessive-far-problem for some of the AIs...it seems to be especially possible to set a special terrain improvemnt as favourite for a leader!)


War

- How aggressive would you rate him or her? And how does that aggressiveness show...a short-tempered character declaring war, when not getting what he/or she wants? Full blown campaigns for idiological reasons? Opportunistic behaviour and going for weak target already in war with third parties? If a war is lost, how strong is the seek for revenge?
- Or is, on the contrary, peace valued a lot?
- Will he/she fight for his own or make use of allies frequently? Are Defensive Pacts an important part of his/her military strategy?
- How confident is he/she about the power and skills of the factions military, how risky and unpredictable does he or she behaves when commanding on the field? (has to been been seen differntly for evaluating overall military size and power on the one hand and estimating odds in single battles)
- Is it likey that a leader commits atrocities (currently only burning down a base)?
- At which point the leader is willing to capitulate?


Diplomacy/Attitude

- How would you describe the general mood of a leader towards others? And is/he or she rather fixed here or behaves erratic (one game nice, next game aggressive etc.)?
- How good have the relations to be, that he/or she is willing to declare war on third party asked by someone else?
- How good have the relations to be, that he/or she is willing to stop trade asked on third party by someone else?
- How is his/her relation vs. other leader with either MORE or LESS power?
- How much does he/she cares about close borders (feels threatened, attitude drops) and how likely is this alone a reason for going to war?
- How much does he/she cares for others having the same religion? OTOH, is there any chance that the leader will adopt to someone elses religion?
- How much does he/she cares for others having the same political ideology? OTOH, is there any chance that the leader will adopt to someone elses ideology?
- How good/bad the attidtude to someone has to be that the leader treats him as friend or foe (giving help, helping in war, no extortition anymore, no attacks vs. the opposite)?
- How good have to be relations to make the leader think about becoming someone's vassal?
- How often does he/she contact someone else...and what will be offered, demanded, traded then mainly?
- How unforgiving or thankful is he/or she if it comes to forgetting past good or bad events and diplomatic happenings?



Eventually it would make sense to do this work leader by leader, so that as soon as enough data is collected for one leader we can mod him or her into Planetfall. This would likely also help to solve some mysterious tags as well, if we can see them in action in the game. Here are the 7 from SMAC which are currently needed in alphabetical order:

Deidre
Lal
Miriam
Morgan
Santiago
Yang
Zhakarov

 

[Pfeffersack]

Another reserved post...

 

[Maniac]

Awesome thread!

I dug up some info which may be useful:

A link:
http://forums.civfanatics.com/showthread.php?t=274966

And an Excel file:

 

[Pfeffersack]

Awesome thread!

I dug up some info which may be useful:

A link:
http://forums.civfanatics.com/showthread.php?t=274966

And an Excel file:

Thanks for the links! The guide you found is a lot more complete then the ones I came up with (though there are still questions open and some strange values just don't seem to fit in its or my own explanations...). I will sum up the content for the second half a bit more, as the tags are more grouped here (for each diplomatic modifer or action, there are usually the same two or three).
The good news is that frequency and effect of each diplomatic action can be modded, the bad that some tags will require more testing to be sure how they work (because here is one of the areas, where Civ4 provides inbuild, but completely unused stuff)

The next step is to form a picture of each Planefall leader. I currently brainstorming here and I would appreciate feedback and ideas of others - just say how you would characterize a special leader. As said above, this can be different from the original SMAC. Look in third post for a plan of mine how to do this.


I will post my own thoughts on Deidre, the first leader on my list, soon.

 

[Maniac]

One way find good values for all the tags might be answering questions about how leader X should behave regarding Y, as with this information it isn't hard to code the behaviour into the tags (though there is no need for everyone to answer every question in line; just use it as guideline in your descriptions, if you like and feel like it might help you to provide feedback)

Yeah, I wouldn't really know the answer to many of these questions, how I would like the leaders to behave.

All fourteen leaders will eventually be included btw. Cha Dawn/Aki probably as Gaian/University leaders.

- Tech Trading ...how interested in acquiring new tech and how willing to share tech in return, especially a new, secret tech no one has yet?

Regarding tech trading in general, I think all leaders should be equally willing to trade techs (or share open borders). By not trading techs or opening borders even with good friends, the AI just shoots itself in the foot. You can see this in unmodded Civ4 where Mansu Masu or whatshisname often performs well, not because his core AI code is different than the others, but simply because he's more willing to trade techs.
I would say though that perhaps the attitude before an AI is willing to trade with you/another AI should be higher than in unmodded Civ. Since, due to conflicting religions and favourite civic, it's very hard to get along with everybody, this would force you to pick sides for succesful gameplay.

- How interested is he/she in secret projects? Or is the entire concept of "secret projects" rather something the leader is against?

A problem here is that Planetfall doesn't have many secret projects yet.
I think it's hard to say though whether or not leaders like to build secret projects in general. Rather it would depend on what the secret project does.

- Generally, how good have the relations to be that a leaders trades something (might be ressoruces, tech, map)? Are there any items the leader will trade even with a worde attitude or not even, when relations are better?

Same answer as with tech trading.

- Relationship to money - saving for domestic purposes or using it for diplomatic/trade activities?

If Morgan Industries was to be made in the Khazad of Planetfall, they would want to hoard money. Not really an opinion about the other leaders. :-s

- How important is espionage (think of military intelligence, security measures vs. sabotage and/or offensive action)?

Angels should favour espionage of course.
I should note her that I'm running around with the following idea: How much you've spent in comparison with your target on espionage, should not affect the mission cost. Instead mission cost is mostly determined by the culture of the target city.

So basically this would mean: Culture = defense against espionage; Espionage = for offensive missions

- How much of the economic capacity is he/she willing to invest in military?

Same as in SMAC I guess: the Hive, Spartans and Believers would be the major military spenders.

War

Morgan would be the most pacifist I guess. Sure, commies are bad for business, but so is war.
Miriam, Deirdre and Lal would be ideological warmongers, for respectively religion, ecology and democracy.
Spartans would be the most rational warmongers. The true realpolitikers. I guess the University would be too.
I guess Yang might not go into full-scale ideological wars, but he might just declare war on you because you've pissed him off, rejected one fof his demands. He demands full obedience after all.

Spartans and Hive (and to a lesser extent University and Morgan) seem like the most likely dogpilers. In other words, the non-ideological warmongers.

Besides dogpiling the weaklings, the Spartans would also oppose the strongest faction if they're not too much outclassed. They want to be top doggie after all. This is a difference with Yang. He would also opportunistically dogpile, but he's more an isolationist, and thus would be less interested in balance of power/countering the strongest.

Lal would always come to the defense of attacked weaker democratic powers. No wars unless they're sanctioned by the UN! Favours defensive pacts.

Gaians and Believers not as much as Lal, but they would also come to the defense of those who share their ideology.

Hive and Believers seem like they might take the most risks in combat. Spartans, despite being warmongers, would not take risks with their units - again, the most calculated warmongers.

Is it likey that a leader commits atrocities (currently only burning down a base)?

I'm not sure if I'm gonna keep base razing at all, so can't really answer that question. Again, base razing seems another one of those "shoot themselves in the foot" tactics.
There's also planet buster use. Yang seems most likely to use those.

- At which point the leader is willing to capitulate?

Morgan and Zakharov would be most willing. Research and commerce can continue even under vassalization. Yang might rather see the world burn around him than give up absolute power. For the others I guess it would depend if you share their religion/favourite civic. But in the end they might surrender to save their people from more suffering.

How is his/her relation vs. other leader with either MORE or LESS power?

Santiago might dislike you simply because you're more powerful. For the others ideology probably matters more.

- How much does he/she cares for others having the same religion? OTOH, is there any chance that the leader will adopt to someone elses religion?

Miriam and Deirdre would be the religious loonies. Others would be flexible in their religion. Wouldn't mind switching.

- How much does he/she cares for others having the same political ideology? OTOH, is there any chance that the leader will adopt to someone elses ideology?

Lal, Deirdre and Miriam would care the most.

 

[Pfeffersack]

Yeah, I wouldn't really know the answer to many of these questions, how I would like the leaders to behave.

As said, I didn't expected that. Your feedback was very helpful and since you gave it for all leaders at once, I realized that my per leader attempt would have probably failed anyway. Even with the BTS leaders as comparison line, the differences between BTS and Planetfall are too big and so is behavuour of the leaders. So I have used you feedback directly to find possible values for all leaders. I will post and Excel file and some explanations in the next post

All fourteen leaders will eventually be included btw. Cha Dawn/Aki probably as Gaian/University leaders.

I'm glad to hear that, but I think we can stick to the orignial seven for the moment. Getting them right is difficult enough, but if we suceed getting in the others will be even easier.

Regarding tech trading in general, I think all leaders should be equally willing to trade techs (or share open borders). By not trading techs or opening borders even with good friends, the AI just shoots itself in the foot. You can see this in unmodded Civ4 where Mansu Masu or whatshisname often performs well, not because his core AI code is different than the others, but simply because he's more willing to trade techs.
I would say though that perhaps the attitude before an AI is willing to trade with you/another AI should be higher than in unmodded Civ. Since, due to conflicting religions and favourite civic, it's very hard to get along with everybody, this would force you to pick sides for succesful gameplay.

Though my understanding is that "a better AI" is not achieved by having all AIs playing the same (leads often collective stupidity and boring games), "optimal" way, I understand your point. Tech trading might be just too powerful and important to cripple any AI here.

A problem here is that Planetfall doesn't have many secret projects yet.
I think it's hard to say though whether or not leaders like to build secret projects in general. Rather it would depend on what the secret project does.

I have the same feeling. Nonetheless I think some slight differences could be fun - e.g. the University should be probably most mad about getting secret projects, while the Spartans could be more tempted to invest in units, to conquer other peoples secret projects...

If Morgan Industries was to be made in the Khazad of Planetfall, they would want to hoard money. Not really an opinion about the other leaders. :-s

I see it the opposite way - Morgan has a lot of money and it is his tool to manipulate the world. Trade is everything. Thats why I would give 30-50% of his money free for trade. Generally, I would up the percentagess a bit compared to BTS, where I always feel that the AIs have not enough money available for trade.

Angels should favour espionage of course.I should note her that I'm running around with the following idea: How much you've spent in comparison with your target on espionage, should not affect the mission cost. Instead mission cost is mostly determined by the culture of the target city.So basically this would mean: Culture = defense against espionage; Espionage = for offensive missions

Agreed, though I think Believers, Hive and Spartans should get a boost as well (because of the way in which they run their societies). University and Gaians are the underperformers here, IMO. That would be for the old model - if culture takes the defensive part, the penalties could be left out.

Same as in SMAC I guess: the Hive, Spartans and Believers would be the major military spenders.

Morgan would be the most pacifist I guess. Sure, commies are bad for business, but so is war.
Miriam, Deirdre and Lal would be ideological warmongers, for respectively religion, ecology and democracy.
Spartans would be the most rational warmongers. The true realpolitikers. I guess the University would be too.
I guess Yang might not go into full-scale ideological wars, but he might just declare war on you because you've pissed him off, rejected one fof his demands. He demands full obedience after all.

Spartans and Hive (and to a lesser extent University and Morgan) seem like the most likely dogpilers. In other words, the non-ideological warmongers.

Besides dogpiling the weaklings, the Spartans would also oppose the strongest faction if they're not too much outclassed. They want to be top doggie after all. This is a difference with Yang. He would also opportunistically dogpile, but he's more an isolationist, and thus would be less interested in balance of power/countering the strongest.

Lal would always come to the defense of attacked weaker democratic powers. No wars unless they're sanctioned by the UN! Favours defensive pacts.

Gaians and Believers not as much as Lal, but they would also come to the defense of those who share their ideology.

Hive and Believers seem like they might take the most risks in combat. Spartans, despite being warmongers, would not take risks with their units - again, the most calculated warmongers.

Sounds resonable. I first didn't liked Deidre as warmonger, but after thinking about it, it feels right - she was characterized as pacifist in SMAC, but in fact it played out as being at war with her all the time, if not running a Green Economy.

I'm not sure if I'm gonna keep base razing at all, so can't really answer that question. Again, base razing seems another one of those "shoot themselves in the foot" tactics. There's also planet buster use. Yang seems most likely to use those.

But only if the AI does it all the time...I think there are situations, where it makes sense. OTOH as you have decreased the destruction happening, when capturing a city, it might be not so wise in Plnetfall. Nonetheless I would preserve the option for flavour, but it should be tied to the Planet Charta or something similar then.

Morgan and Zakharov would be most willing. Research and commerce can continue even under vassalization. Yang might rather see the world burn around him than give up absolute power. For the others I guess it would depend if you share their religion/favourite civic. But in the end they might surrender to save their people from more suffering.

Santiago might dislike you simply because you're more powerful. For the others ideology probably matters more.

Miriam and Deirdre would be the religious loonies. Others would be flexible in their religion. Wouldn't mind switching.

Lal, Deirdre and Miriam would care the most.

I agree.

 

[Pfeffersack]

Attached is my first attempt of completely reworking the leaders - in theory (so I haven't started to change the leaderhead.xml yet). It's a file about the values of the BTS leaders in 3.13 and I have simply added the 7 from Planetfall underneath.

What I haven't touched so far are the flavour entries and the complete 5th and 6th section of Part II - I don't want to fiddle with the memories, as long as I don't really understand and I'm not sure if the the intervalls in which the AIs ask for something should be changed.

If the values are acceptable or after doing necessary corrections, I would be willing to code them into the leaderhead.xml as well.

Imporant changes:

- Clearer distinction between warmongers and pacifists (regarding respect, weight for peace)
- completely reworked likelihood for the different types of wars, e.g. the "fanatic" leaders will constantly seek for their ideological enemy, but don't have a lot of interest in dogpile wars any more
- more gold for trade (how much more depends on type leader)
- streamlined (tech) trading interest - no more "Tokugawas" and "Mansa", but overall better relations needed
- Leaders show more personality when demands are refused (will they go to war at all and how sneaky they will be?)
- Religion and Civics now matter for the leaders you would expect them to matter
- If leaders are opportunitic or fanatic should now affect their rection on diplomatic proposals more (e.g. when will they enter a war vs. a third person?)
- their active diplomacy should reflect their personality a lot more - expect tribute demands from the bad guys (and girls ), conversion attempts from religious leaders and so on
- also certain leaders now react stronger (both postive or negative) attitude wise, when you behave at their will or step on their toes regarding (for them) "touchy subjects"
- the important group of values for DoW of the AIs on certain relation levels (mainly known as what determines if civs can backstab at "Pleasent") has been reworked - the more a leader cares about your actions in an ideological way, the less he or she is willing to backstab. But watch out for the opportunist ones...

 

[Maniac]

Though my understanding is that "a better AI" is not achieved by having all AIs playing the same (leads often collective stupidity and boring games), "optimal" way, I understand your point. Tech trading might be just too powerful and important to cripple any AI here.

I should note btw, as opposed to my earlier post, this does not count for Open Borders. The Autarky civic means not having open borders doesn't have to equal 'foot-shooting'. So Yang could be very hard to get open borders with, while Morgan and Lal would sign the agreement with everyone.

I have the same feeling. Nonetheless I think some slight differences could be fun - e.g. the University should be probably most mad about getting secret projects, while the Spartans could be more tempted to invest in units, to conquer other peoples secret projects...

Having looked at your Excel file, I'd suggest dropping Deirdre to 20, and raising Yang to 30 - he's an industrial type after all.

I see it the opposite way - Morgan has a lot of money and it is his tool to manipulate the world. Trade is everything. Thats why I would give 30-50% of his money free for trade. Generally, I would up the percentagess a bit compared to BTS, where I always feel that the AIs have not enough money available for trade.

I'll refrain from comment on Morgan until there's a more concrete image what gameplay benefits to give him.
I definitely agree the AI doesn't have enough money for trade. The basic problem here though is that they simply have little credits in their reserves, often less than 100. 50% of nothing is still nothing. So it would require some SDK work to get the AI to keep a larger credit reserve.

Agreed, though I think Believers, Hive and Spartans should get a boost as well (because of the way in which they run their societies). University and Gaians are the underperformers here, IMO. That would be for the old model - if culture takes the defensive part, the penalties could be left out.

With culture as espionage defense, it seems more fitting to me that the Believers would focus on culture, but not on espionage per se.

Nonetheless I would preserve the option for flavour, but it should be tied to the Planet Charta or something similar then.

Yeah, I was thinking you could only raze bases (and build planet busters) if you haven't signed the UN Charter (I could make that a seperate 'civic' category). I could also give some advantages to certain leaders for base razing. Eg the Pirates get much more money from it, and if Cha Dawn did it the Flowering Counter would increase.

I'd like base razing to take several turns though (depending on population size). Otherwise a single mistake has too high consequences, and combat will be too focused on base defenses.

What I haven't touched so far are the flavour entries and the complete 5th and 6th section of Part II - I don't want to fiddle with the memories, as long as I don't really understand and I'm not sure if the the intervalls in which the AIs ask for something should be changed.

Yeah, regarding AIs asking for something... I'm thinking refusing such a demand should not carry any negative consequences. For three reasons:
1) AIs never make each other such demands, so it ends up merely being a general diplo penalty for the human player.
2) That may just be my personal experience, but I never give in to such demands anyway. They just annoy me.
3) Instead of giving a penalty for for instance refusing to change religions, you could simply give the leader a higher penalty for not having the same state religion. Same effect in the long term.

While those demands annoy me, others would probably prefer to keep them - even if they didn't have any diplomatic effect - for the chance to show off SMAC's diplomatic flavour dialogue.

If the values are acceptable or after doing necessary corrections, I would be willing to code them into the leaderhead.xml as well.

The effects of AI changes can never by fully predicted, so I'd put them in the XML and then see if they work out as intended!

Here are some comments on what you have already written in the Excel file:

iBaseAttitude: I'm wondering if - like tech trading - this isn't one of the cases where everyone should have the same value??

iBuildUnitProb: perhaps drop the values here a little?? I mean, in unmodded Civ there are only a handful who have value 40, but in your Excel there are no less than three of the seven! In my observation, the AI often builds many military units already - they just don't do anything with them...

iBetterRankDifferenceAttitudeChange: I assume this value affects relation towards players weaker than the leader in question. I noticed you gave Santiago a -1 here. I'm rather thinking that Santiago would not look automatically have worse relations with weaker players. It's just, she would be more likely to declare war on weak players, regardless of how she likes them. ,)

 

[Pfeffersack]

I'll refrain from comment on Morgan until there's a more concrete image what gameplay benefits to give him.
I definitely agree the AI doesn't have enough money for trade. The basic problem here though is that they simply have little credits in their reserves, often less than 100. 50% of nothing is still nothing. So it would require some SDK work to get the AI to keep a larger credit reserve.

My experience is different, especially for Planetfall - it is not so uncommon that the AIs have up to a few hundred credits for trading (what can mean up to several thousand in the treasury, if the trade percent is only 5% or 10% - which is the setting for most BTS leaders) However, it might be that my experience is cuased by playing on Emperor difficulty mainly - in normal BTS (where I play either Prince or Monarch), money is shorter - at least up to the industrial age, where the AIs treasuries seem to grow.

I'd like base razing to take several turns though (depending on population size). Otherwise a single mistake has too high consequences, and combat will be too focused on base defenses.

That sounds great. It reminds me a bit of the Age of Wonders series, where you have two options to get rid of a captured city - either plain destruction (always within 1 turn, but no plunder taken) or looting, which tooks one turn per city size (up to 4) and gives you Gold.

Yeah, regarding AIs asking for something... I'm thinking refusing such a demand should not carry any negative consequences. For three reasons:
1) AIs never make each other such demands, so it ends up merely being a general diplo penalty for the human player.
2) That may just be my personal experience, but I never give in to such demands anyway. They just annoy me.
3) Instead of giving a penalty for for instance refusing to change religions, you could simply give the leader a higher penalty for not having the same state religion. Same effect in the long term.

While those demands annoy me, others would probably prefer to keep them - even if they didn't have any diplomatic effect - for the chance to show off SMAC's diplomatic flavour dialogue.

Hmm, I'm not so clearly apposed against the AI demands. Surely, 1) annoys me to no end as welland sadly there seems to be no way to change this. For 2) - I actually give in to those demands occasionally, usually if I can afforf to do so and if keeping up relations with the demanding AI is important for me. 3) makes kind of sense and if also more "fair", because you get the effect everytime (and not as per turn chance of a demand)
I would keep them for flavour reasons, but then it's kind of strange that no effect is attached. What I have already done though (and what surely could be taken further), is reducing the negative reaction of some leaders - why not remove the negative reaction for all leaders, except the ones really mad about the item? This would mean in case of religion e.g. only Miriam and maybe Deidre would like you -1 less if you refuse, while you get away without penalty from the others.

iBaseAttitude: I'm wondering if - like tech trading - this isn't one of the cases where everyone should have the same value??

My intention was to make the warlike leaders are bit more angry in general, while e.g. Morgan mood would be rather positive vs. anyone in the beginning - everyone could be a trading partner. But it is not impossible that this causes strange sideeffects (e.g. the same to AI leaders never getting along with each other just because on is "angry" and the other "friendly")

iBuildUnitProb: perhaps drop the values here a little?? I mean, in unmodded Civ there are only a handful who have value 40, but in your Excel there are no less than three of the seven! In my observation, the AI often builds many military units already - they just don't do anything with them...

Thinking about it, you might be right that I have gone overboard with this.

iBetterRankDifferenceAttitudeChange: I assume this value affects relation towards players weaker than the leader in question. I noticed you gave Santiago a -1 here. I'm rather thinking that Santiago would not look automatically have worse relations with weaker players. It's just, she would be more likely to declare war on weak players, regardless of how she likes them. ,)

No, it's the exact opposite (I double-checked again) - the relation towards someone stronger. I gave the -1 because of your hint on Santiago challenging the top dog. Relations towards weaker leaders are determined by "iWorseRankDifferenceAttitudeChange".

But you are right; I have mixed up "worse relations" and "likely to dogpile" here (I attached even -3 for Santiago here). Probably no need to double-punish weaker factions here.

 

[Maniac]

My experience is different, especially for Planetfall - it is not so uncommon that the AIs have up to a few hundred credits for trading

I was more thinking the AI should have like 1000 credits up for trade, when the techs being trades also start costing 1000s of beakers.

I would keep them for flavour reasons, but then it's kind of strange that no effect is attached.

Here's another alternative: the AI should actually *reward* you for following their suggestions. "If you switch to Voice of Planet, I will give you knowledge of Centauri Empathy."
Would require SDK modifications though.

 

[Pfeffersack]

Here's another alternative: the AI should actually *reward* you for following their suggestions. "If you switch to Voice of Planet, I will give you knowledge of Centauri Empathy."
Would require SDK modifications though.

Nice idea, I see only the problem of exploiting the AI here - what would happen if you switch to only reap the gift? It might work, if on the other hand the penalty for refusing is kept and if switching a way later would have serious negative consequences status wise (e.g. high penalty for just having a different religion; there is also a parameter, which regulates how fast the attitude drops for this - I think those "Divisor" thing, though there is that said anomaly with the different religion tags)



A suggestion - if you start to implement changes to the leaderheads and decide to use different values then the ones I suggested in my file, could you update it please?
Could be helpful for testing purposes, as then everyone can theorize about what could have caused behaviour X and where further improvemnts could be made eventually. Also, it is likely easier to maintain such a file then to build it up from scratch again, if one day the SMAX leaders get added and comparison values might be helpful.

 

[Maniac]

Nice idea, I see only the problem of exploiting the AI here - what would happen if you switch to only reap the gift?

Switching civics or religion still costs you a turn of anarchy. So the trick is making the gift more or less the same value as a turn of anarchy costs you. That way it's still only a good idea to switch if the new civic/religion is better than your current one.

In fact, why not make it so that a leader does not give you any gifts, but that switching civics/religions on the request of a leader does not cost you a turn of anarchy? It would be much easier to code - no need to write an AI to judge the value of a gift.

A suggestion - if you start to implement changes to the leaderheads and decide to use different values then the ones I suggested in my file, could you update it please?

Sure. Just to be sure, does that mean you're not gonna put the values in XML?

 

[Pfeffersack]

In fact, why not make it so that a leader does not give you any gifts, but that switching civics/religions on the request of a leader does not cost you a turn of anarchy? It would be much easier to code - no need to write an AI to judge the value of a gift.

A very simple and elegant solution The anarchy is sometimes what makes me refuse a switch (only to delay it a few turns to be able to switch as second civic at once), so I think this would add to the decision. Moreover, switching out again would still cost you a turn anarchy.

Sure. Just to be sure, does that mean you're not gonna put the values in XML?

I originally planned to do so, but I thought you would prefer to do this yourself, when reading your reply to my values file. I would be still willing and able to take that job, just let me know if you wish any further corrections.

I also don't feel to comfortable with adding lines to the leaderhead.xml (the improvement weight entries at the end come to my mind, as they could help to make certain leaders go in the Terraform or Hybrid direction, but there are no entries in Planetfall leaderheads yet), so I would prefer to leave that part to you.

 

[Pfeffersack]

Here is the list of corrections I have made so far to my original file based on your feedback:

iWonderConstructRand

DEIDRE 20 (was 30)
YANG 30 (was 20)


iBaseAttitude

MORGAN 0 (had 2)
SANTIAGO 0 ( had -1)
YANG 0 (had -1)
ZHAK 0 (had 1)


iBuildUnitProb

DEIDRE 25 (was 30)
LAL 25 (was 30)
MIRIAM 25 (was 40)
MORGAN 15 (was 20)
SANTIAGO 30 (was 40)
YANG 30 (was 40)
ZHAKAROV 20 (was 30)


iWorseRankDifferenceAttitudeChange

MORGAN 0 (was -1)
SANTIAGO 0 (was -3)
YANG 0 (was -3)
ZHAK 0 (was -1)


iOpenBordersRefuseAttitudeThreshold

LAL furious (was annoyed)
MORGAN furious (was annoyed)
YANG cautious (was annoyed)

 

[Maniac]

I originally planned to do so, but I thought you would prefer to do this yourself, when reading your reply to my values file. I would be still willing and able to take that job, just let me know if you wish any further corrections.

I'd very much appreciate it if you would put those values in the XML.
Then I can for instance spend time earlier to put the civic/religion switch idea in the SDK.

I also don't feel to comfortable with adding lines to the leaderhead.xml (the improvement weight entries at the end come to my mind, as they could help to make certain leaders go in the Terraform or Hybrid direction, but there are no entries in Planetfall leaderheads yet), so I would prefer to leave that part to you.

Yeah, no need to touch terrain improvement preferences.

 

[Pfeffersack]

I have completed editing the leaderheads.xml now. I have attached two versions of it - v1 is exactly what we have discussed so far, while v2 incorporates some changes I have made after doing some playtesting (a fun game - more AI wars based on ideology, building up good relations is more essentially now, leaders act differently in regard to what they demand and are willing to do, more trading possible) - some things just don't played out as intended (see explanation of changes below).

Attached is also again the table with both the v1 and the changes for v2 (red/green marked values).

Changed in v2

1. Everyone has 80% of Gold and GpT availabe for trade. The values in BTS are far too low for my taste and just doubling from 5 to 10 or ocassionally 20 just isn't enough. The idea behind those values is to avoid tech brokering and bankrupting the AI. But I don't think this is such a big issue, that all AIs should keep 70% to 95% of their Gold away from trading. They get a discount on unit upgrades, they don't give a lot of Gold for tech anyway and you can't trade in GpT for tech. 20% a reserve should be enough and the gains from additional tech trades should help them (money just cannot act as currency in Civ4, if there is not at least enough circulating to pay the beaker differences between techs)
I have choosed a flat value for the moment (also more fair because of the tech-trade impact); we can bring back individual differences anytime we are sure about how single leaders should behave in financial matters.

2. Some minor changes to impact of sharing favourite civic and same/different religion - if we think of religions in Planetfall in a broader sense("idealogies", "movements"), they should matter for most leaders (except maybe Santiago and Morgan). Even Yang and Zhakarov could be imagined as following religions liked Ascetic Virtues or the Consciouness. So I gave them slightly higher values. Lal on the other hand is more plotical focussed and could be imagined as rather tolerant towards other religions. I also came to the conclusion that both Deidre and Miriam need some emphasize on the right civics as well - they are the main ideological competitors regarding Hybrid/Terraformed and while both directions are attached to religions, they found on certain civics.

3. Lal and Morgan caring more for open Borders and/or Trade. I think it adds some spice and flavour to the usual mixture of civic and religion influence, if some leaders are especially interested in stable economic relations. For Lal it is also a way to fill that desire for Democracy with some more life.

4. Several changes were necessary in regard to the tresholds values for certain diplomatic actions of the AI. I missunderstood the impact of those tresholds first - I thought "PLEASED" would mean that a leader would be willing to do then something when his/her mood is pleased - but it means in fact that he/she needs to be "FRIENDLY" This lead to some too harsh restrictions.

5. I incorporated the spirit of the changes under 2) and 3) also into the contact chances and diplomatic penalties/boni for certain actions - especially the fanatic leaders show now more activity (demand tribute more often, which is however restricted by the changes under 4) to those who they don't like and both ask to change civics and religion more often). Yang can be now better appeased by meeting his demands and Morgan/Lal show their preference for free trade.

I recommend v2 of course, but if there is anything you don't like, I will change it back again


Note: XML-files have to be renamed (remove "-v1" or "-2") before use

 

[Maniac]

Thanks - downloaded. I'll check i out!

Edit: This post has been thrown into the recycling tanks!

 

[Pfeffersack]

While playtesting the file I noticed also some things which I could not adress:

- While the combination of Voice of Planet and Terraformed isn't allowed, it is possible to run Edenism under Hybrid (especially an AI controlled Miriam is prone to do that in games she doesn't switch to Terraformed) - is that intended?

- I noticed that the religios flavour section in the leaderhead.xml isn't used for Planetfall yet. My suggestion would be to attach each religion to one leader (Morgan and Santiago left without a preference):

Deidre = VoP
Lal = Homo Superior
Miriam = Edenism
Yang = Ascetic Virtues
Zhak = The Consciousness

Jugding from what "favourite civic" does - a moderat leaning towards the civic - I think that would be interesting for religion as well. It adds to flavour (without fixing things completely), if a religious Yang or Zhakarov follow religions which fit their overall ideology. Also, the "Choose Religion" setting in the game setup would work then.

- Another limit is reached when it comes to modeling the effect of different or (better said) opposed civics - there is no such entry like we have it for religions. Planetfall has it for Hybrid/Terraformed, but SMAC went much further with that. It just does not feel right that Yang and Lal are best friends and that Morgan hardly cares about my Planned Economy. I tried to counter that to some extend by increasing the likelihood of civic switch demands and the attached penalties for refusing, but that of course only affects relations to the human player. I would suggest to add that kind of extra-tension also to pair likes Democracy/Police State and Planned/Free Market.

- That would be probably hardest to code...a thing I miss from SMAC is that AIs sometimes offered you a "compensation for your expenses" (credits, tech and/or an alliance), if you entered a war on their behave. I find myself declining most war requests in Civ4/Planetfall (except the ones from very good friends) comapred to SMAC, just because it is often not worth to declare war someone you have rather neutral relations with, if someone asked you with the same or only litter better relations - you lose more on one side (and probably third parties as "friends" as well) than what you get on the other. This would be much more interesting, if AIs demanding going to war - while having annoyed or cautious relations with me - would offer something. It would be also only fair, because the human player needs to pay the AI nearly always for going to war, even good friends (only exception seems to be if they were already at war with the target earlier)